1. Field of the Invention
The present invention relates a coordinates input apparatus and, more particularly, to a coordinates input apparatus for detecting the coordinates of an instruction point on a vibration transmitting plate on the basis of the vibration transmission time on the vibration transmitting plate.
2. Related Background Art
A coordinates input apparatus, particularly, an apparatus for inputting an instruction point onto a tablet by using an input pen is an input apparatus which is inevitable for a hand writing recognizing system for recognizing hand-written characters or figures in an on-line manner.
In the coordinates input apparatus, the instruction point is frequently erroneously detected due to the causes such as noises or the like. Such an erroneous detection point causes the hand writing shape of the input character or figure to be deformed. Therefore, such an erroneous detection point becomes one of the causes of the erroneous recognition in the hand writing recognizing system. Consequently, a technique to eliminate such an erroneous detection point is very important.
Hitherto, in the above input apparatus, the following systems for eliminating the erroneous detection point have generally been used.
The coordinates input apparatus samples the coordinates of the instruction point every predetermined time (hereinafter, referred to as a sampling time). In general, on the assumption that a distance d between two adjacent sampling points of a hand-written straight line or curve does not exceed a predetermined value d.sub.max, when the distance d exceeds the value d.sub.max, the occurrence of the erroneous detection is determined and the instruction points which were sampled after such a determination are regarded as the erroneous detection points and eliminated.
For instance, as shown in FIG. 9, when there are instruction points P.sub.1, P.sub.2, . . . , P.sub.n which are time sequentially sampled and an erroneous detection point P.sub.4 exists, ##EQU1## (where, d.sub.i,i+l denotes a distance between two points P.sub.i and P.sub.i+l and x.sub.i and y.sub.i indicate coordinates values of P.sub.i). Therefore, P.sub.4 is eliminated as an erroneous detection point.
The distance d can be obtained by a method of using the distance of the straight line between two points as shown in the equation (1). In addition to this method, it is also possible to use a method whereby only the components of x and y axes are fetched and the d.sub.max is individually discriminated as shown in the following equations (2) and (3). EQU xd.sub.i,i+l =.vertline.x.sub.i -x.sub.i+l .vertline.&gt;xd.sub.max . . . ( 2) EQU yd.sub.i,i+l =.vertline.y.sub.i -y.sub.i+l .vertline.&gt;yd.sub.max . . . ( 3)
(where, xd.sub.i,i+l and yd.sub.i,i+l represent components of x and y axes of the distance between two points P.sub.i and P.sub.i+l)
However, in the foregoing conventional example, when the hand writing is executed at a speed exceeding the d.sub. max for the sampling time .DELTA.t, the correct instruction points other than the erroneous detection points due to the noises or the like are also eliminated. That is, a hand writing velocity v.sub.w cannot exceed ##EQU2## For instance, assuming that .DELTA.t=10 msec and d.sub.max =1 mm, ##EQU3## This value is fairly smaller than the ordinary hand writing speed of tens of cm/sec. This means that according to such a system, characters, images, or the like cannot be input by the ordinary hand writing speed.
It is proper to increase the value of d.sub.max to ease the limitation of the hand writing speed. However, the erroneous detection points cannot be sufficiently eliminated by such a method. Therefore, there is a drawback such that if the user tries to eliminate the sufficiently significant erroneous detection points by the conventional technique, the hand writing speed is remarkably limited.
On the other hand, according to the conventional technique, it is impossible to discriminate whether the first sampling point P.sub.1 is the erroneous detection point or not. Therefore, when the first point is the erroneous detection point, there is also a drawback such that not only this point cannot be eliminated but also the correct second and subsequent sampling points P.sub.2 to P.sub.n after the first erroneous sampling point are erroneously eliminated as the erroneous detection points.